“Uncanny Valley 2” is a study that examines children’s beliefs and feelings about a collection of real-world robots based on a viewing of an 8-second video of that robot.
## [1] "2017-02-18"
## [1] "2018-02-08"
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.0 9.0 14.0 18.1 22.5 54.0
## [1] 120
| 3-6 |
15 |
21 |
36 |
| 6-9 |
18 |
22 |
40 |
| 9-12 |
17 |
17 |
34 |
| >12 |
4 |
6 |
10 |
| Sum |
54 |
66 |
120 |
| 3.413 |
5.408 |
7.772 |
7.914 |
9.789 |
17.46 |
## [1] 120
## [1] 117
## [1] 113
Descriptives

Analysis of Questions
Confirmatory Factor Analysis
## lavaan (0.5-23.1097) converged normally after 82 iterations
##
## Number of observations 288
##
## Estimator ML
## Minimum Function Test Statistic 45.797
## Degrees of freedom 17
## P-value (Chi-square) 0.000
##
## Model test baseline model:
##
## Minimum Function Test Statistic 631.689
## Degrees of freedom 28
## P-value 0.000
##
## User model versus baseline model:
##
## Comparative Fit Index (CFI) 0.952
## Tucker-Lewis Index (TLI) 0.921
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -3279.481
## Loglikelihood unrestricted model (H1) -3256.583
##
## Number of free parameters 19
## Akaike (AIC) 6596.962
## Bayesian (BIC) 6666.558
## Sample-size adjusted Bayesian (BIC) 6606.307
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.077
## 90 Percent Confidence Interval 0.050 0.104
## P-value RMSEA <= 0.05 0.048
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.053
##
## Parameter Estimates:
##
## Information Expected
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## agency =~
## choose 1.000
## think 1.032 0.107 9.663 0.000
## moral 0.787 0.096 8.182 0.000
## exp =~
## scared 1.000
## pain 1.043 0.131 7.969 0.000
## hungry 1.580 0.185 8.523 0.000
## uv =~
## creepy 1.000
## weird 2.995 2.986 1.003 0.316
##
## Covariances:
## Estimate Std.Err z-value P(>|z|)
## agency ~~
## exp 0.381 0.064 5.911 0.000
## uv -0.061 0.071 -0.860 0.390
## exp ~~
## uv -0.018 0.024 -0.736 0.462
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .choose 0.751 0.093 8.109 0.000
## .think 0.629 0.088 7.120 0.000
## .moral 0.997 0.097 10.266 0.000
## .scared 0.743 0.072 10.348 0.000
## .pain 0.608 0.063 9.692 0.000
## .hungry 0.374 0.083 4.490 0.000
## .creepy 1.120 0.256 4.378 0.000
## .weird -1.021 2.138 -0.478 0.633
## agency 0.808 0.133 6.085 0.000
## exp 0.380 0.079 4.807 0.000
## uv 0.242 0.247 0.979 0.328
## $lambda
## agency exp uv
## choose 0.720 0.000 0.000
## think 0.760 0.000 0.000
## moral 0.578 0.000 0.000
## scared 0.000 0.582 0.000
## pain 0.000 0.636 0.000
## hungry 0.000 0.847 0.000
## creepy 0.000 0.000 0.421
## weird 0.000 0.000 1.375
##
## $theta
## choose think moral scared pain hungry creepy weird
## choose 0.482
## think 0.000 0.422
## moral 0.000 0.000 0.666
## scared 0.000 0.000 0.000 0.662
## pain 0.000 0.000 0.000 0.000 0.595
## hungry 0.000 0.000 0.000 0.000 0.000 0.283
## creepy 0.000 0.000 0.000 0.000 0.000 0.000 0.823
## weird 0.000 0.000 0.000 0.000 0.000 0.000 0.000 -0.891
##
## $psi
## agency exp uv
## agency 1.000
## exp 0.688 1.000
## uv -0.139 -0.059 1.000
Partial Correlations

A Priori Variables





##
## Pearson's product-moment correlation
##
## data: RBI$exp.c and RBI$agency.c
## t = 8.8, df = 290, p-value <0.0000000000000002
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.3642 0.5469
## sample estimates:
## cor
## 0.4604
##
## Call:
## lm(formula = uv.c ~ exp.c + agency.c + Sex + Age, data = RBI)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.253 -0.829 -0.371 0.631 2.438
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.01318 0.27410 -0.05 0.962
## exp.c 0.12304 0.07029 1.75 0.081 .
## agency.c -0.16103 0.06819 -2.36 0.019 *
## Sex 0.05185 0.11803 0.44 0.661
## Age -0.00832 0.02346 -0.35 0.723
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.995 on 283 degrees of freedom
## Multiple R-squared: 0.0239, Adjusted R-squared: 0.0101
## F-statistic: 1.73 on 4 and 283 DF, p-value: 0.143
##
## Call:
## lm(formula = uv.c ~ exp.c + agency.c + Sex + Age, data = RBI)
##
## Standardized Coefficients::
## (Intercept) exp.c agency.c Sex Age
## 0.00000 0.12304 -0.16103 0.02588 -0.02412
K-means clustering of a priori variables

## [1] 95.2
##
## Call:
## lm(formula = uv ~ cluster.name, data = km)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.000 -0.894 -0.394 0.606 2.328
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.9394 0.1733 11.19 <0.0000000000000002 ***
## cluster.nameHALE -0.2678 0.2117 -1.26 0.21
## cluster.nameLAHE 0.0606 0.2583 0.23 0.81
## cluster.nameLALE -0.0450 0.1902 -0.24 0.81
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.995 on 284 degrees of freedom
## Multiple R-squared: 0.0116, Adjusted R-squared: 0.00116
## F-statistic: 1.11 on 3 and 284 DF, p-value: 0.345
##
## Call:
## lm(formula = uv ~ cluster.name, data = km)
##
## Standardized Coefficients::
## (Intercept) cluster.nameHALE cluster.nameLAHE cluster.nameLALE
## 0.00000 -0.11379 0.01777 -0.02247
Distribution of robots among k-means clusters
##
## Call:
## lm(formula = uv ~ agency.c + exp.c + robot.group + Sex + Age,
## data = RBI)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.342 -0.772 -0.342 0.722 2.413
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.6089901 0.3397258 4.74 0.0000035 ***
## agency.c -0.1494664 0.0671771 -2.22 0.027 *
## exp.c 0.1194660 0.0692261 1.73 0.085 .
## robot.grouphuman-like 0.4574003 0.2484684 1.84 0.067 .
## robot.grouprobotic 0.0519473 0.2325755 0.22 0.823
## Sex 0.0645061 0.1161184 0.56 0.579
## Age -0.0000625 0.0233074 0.00 0.998
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.978 on 281 degrees of freedom
## Multiple R-squared: 0.0554, Adjusted R-squared: 0.0353
## F-statistic: 2.75 on 6 and 281 DF, p-value: 0.013
##
## Call:
## lm(formula = uv ~ agency.c + exp.c + robot.group + Sex + Age,
## data = RBI)
##
## Standardized Coefficients::
## (Intercept) agency.c exp.c
## 0.0000000 -0.1500786 0.1199553
## robot.grouphuman-like robot.grouprobotic Sex
## 0.1992176 0.0243625 0.0323287
## Age
## -0.0001818
Data-driven aggregates
K-means clustering of a data-driven components

## [1] 250.7
Distribution of robots among k-means clusters (PCA)
##
## 1 2 3
## kf 1 0 0
## pepper 1 0 0
## sofia 1 0 0
## atlas 0 1 0
## nao 0 1 0
## spot 0 1 0
## stan 0 1 0
## tapia 0 1 0
## actroid 0 0 1
## festo 0 0 1
## kb 0 0 1
##
## Call:
## lm(formula = uv ~ cluster, data = km)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.111 -0.750 -0.419 0.623 2.323
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.6769 0.1231 13.62 <0.0000000000000002 ***
## cluster2 0.2419 0.1468 1.65 0.100
## cluster3 0.4342 0.2272 1.91 0.057 .
## cluster4 0.0731 0.1965 0.37 0.710
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.992 on 284 degrees of freedom
## Multiple R-squared: 0.0173, Adjusted R-squared: 0.00691
## F-statistic: 1.67 on 3 and 284 DF, p-value: 0.175
##
## Call:
## lm(formula = uv ~ cluster, data = km)
##
## Standardized Coefficients::
## (Intercept) cluster2 cluster3 cluster4
## 0.00000 0.12137 0.12730 0.02594
Imputed Data
Analysis of Questions
Exploratory/Confirmatory Factor Analysis
## ** WARNING ** lavaan (0.5-23.1097) did NOT converge after 747 iterations
## ** WARNING ** Estimates below are most likely unreliable
##
## Number of observations 1243
##
## Estimator ML
## Minimum Function Test Statistic NA
## Degrees of freedom NA
## P-value NA
##
## Parameter Estimates:
##
## Information Expected
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## agency =~
## choose 1.000
## think 0.966 NA
## moral 0.933 NA
## exp =~
## scared 1.000
## pain 0.943 NA
## hungry 1.104 NA
## uv =~
## creepy 1.000
## weird 0.000 NA
##
## Covariances:
## Estimate Std.Err z-value P(>|z|)
## agency ~~
## exp 0.705 NA
## uv 0.220 NA
## exp ~~
## uv 0.233 NA
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .choose 0.505 NA
## .think 0.432 NA
## .moral 0.515 NA
## .scared 0.317 NA
## .pain 0.326 NA
## .hungry 0.251 NA
## .creepy -13276.450 NA
## .weird 1.153 NA
## agency 0.922 NA
## exp 0.712 NA
## uv 13277.698 NA
## $lambda
## agency exp uv
## choose 0.804 0.000 0.000
## think 0.816 0.000 0.000
## moral 0.780 0.000 0.000
## scared 0.000 0.832 0.000
## pain 0.000 0.813 0.000
## hungry 0.000 0.881 0.000
## creepy 0.000 0.000 103.156
## weird 0.000 0.000 0.005
##
## $theta
## choose think moral scared pain hungry
## choose 0.354
## think 0.000 0.334
## moral 0.000 0.000 0.391
## scared 0.000 0.000 0.000 0.308
## pain 0.000 0.000 0.000 0.000 0.339
## hungry 0.000 0.000 0.000 0.000 0.000 0.224
## creepy 0.000 0.000 0.000 0.000 0.000 0.000
## weird 0.000 0.000 0.000 0.000 0.000 0.000
## creepy weird
## choose
## think
## moral
## scared
## pain
## hungry
## creepy -10640.127
## weird 0.000 1.000
##
## $psi
## agency exp uv
## agency 1.000
## exp 0.870 1.000
## uv 0.002 0.002 1.000
Partial Correlations

A Priori Variables





##
## Pearson's product-moment correlation
##
## data: RBI.imp$exp.c and RBI.imp$agency.c
## t = 39, df = 1200, p-value <0.0000000000000002
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.7136 0.7640
## sample estimates:
## cor
## 0.7398
##
## Call:
## lm(formula = uv.c ~ exp.c + agency.c + Sex + Age, data = RBI.imp)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.426 -0.805 -0.357 0.647 2.805
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.0603 0.0949 -0.64 0.53
## exp.c 0.2333 0.0424 5.51 0.000000045 ***
## agency.c -0.1689 0.0423 -4.00 0.000067806 ***
## SexMale -0.0652 0.0565 -1.16 0.25
## Age 0.0120 0.0105 1.14 0.25
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.989 on 1238 degrees of freedom
## Multiple R-squared: 0.0256, Adjusted R-squared: 0.0225
## F-statistic: 8.14 on 4 and 1238 DF, p-value: 0.00000177
K-means clustering of a priori variables

## [1] 370.9
Distribution of robots among k-means clusters
##
## 1 2 3
## festo 1 0 0
## atlas 0 1 0
## kb 0 1 0
## kf 0 1 0
## spot 0 1 0
## actroid 0 0 1
## nao 0 0 1
## pepper 0 0 1
## sofia 0 0 1
## stan 0 0 1
## tapia 0 0 1
##
## Call:
## lm(formula = uv ~ cluster, data = km)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.943 -0.793 -0.293 0.707 2.467
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.7927 0.0315 57.00 <0.0000000000000002 ***
## cluster2 0.1502 0.0785 1.91 0.0558 .
## cluster3 -0.2602 0.0828 -3.14 0.0017 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.951 on 1240 degrees of freedom
## Multiple R-squared: 0.0125, Adjusted R-squared: 0.0109
## F-statistic: 7.86 on 2 and 1240 DF, p-value: 0.000404
##
## Call:
## lm(formula = uv ~ cluster, data = km)
##
## Standardized Coefficients::
## (Intercept) cluster2 cluster3
## 0.00000 0.05465 -0.08969
##
## Call:
## lm(formula = uv ~ agency.c + exp.c + robot.group + Sex + Age,
## data = RBI)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.481 -0.751 -0.359 0.615 2.627
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.3467 0.1054 12.77 < 0.0000000000000002 ***
## agency.c -0.1786 0.0396 -4.51 0.00000714900446 ***
## exp.c 0.2427 0.0398 6.10 0.00000000139313 ***
## robot.grouphuman-like 0.6416 0.0872 7.36 0.00000000000034 ***
## robot.grouprobotic 0.4041 0.0700 5.77 0.00000000975618 ***
## SexMale -0.0614 0.0528 -1.16 0.25
## Age 0.0118 0.0098 1.21 0.23
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.925 on 1236 degrees of freedom
## Multiple R-squared: 0.0682, Adjusted R-squared: 0.0636
## F-statistic: 15.1 on 6 and 1236 DF, p-value: <0.0000000000000002
##
## Call:
## lm(formula = uv ~ agency.c + exp.c + robot.group + Sex + Age,
## data = RBI)
##
## Standardized Coefficients::
## (Intercept) agency.c exp.c
## 0.00000 -0.18678 0.25388
## robot.grouphuman-like robot.grouprobotic SexMale
## 0.25892 0.20339 -0.03202
## Age
## 0.03668
Data-driven aggregates
Principal Components Analysis

## PC1 PC2 PC3 PC4 PC5 PC6 PC7
## choose 0.1391 0.14729 0.19039 0.385614 0.08347 0.02297 0.03289
## feel 0.1507 0.10041 0.13581 0.062619 0.16262 0.02872 0.38248
## hungry 0.1510 0.07863 0.02580 0.051288 0.27548 0.20673 0.25560
## moral 0.1390 0.20082 0.21480 0.069632 0.10073 0.17646 0.23018
## pain 0.1362 0.24398 0.03446 0.068407 0.04717 0.33947 0.02926
## scared 0.1419 0.17948 0.11852 0.009152 0.28702 0.21210 0.05032
## think 0.1423 0.04938 0.28021 0.353288 0.04352 0.01355 0.01927
## Importance of components:
## PC1 PC2 PC3 PC4 PC5 PC6 PC7
## Standard deviation 2.195 0.8335 0.6217 0.5899 0.5341 0.5039 0.4622
## Proportion of Variance 0.688 0.0993 0.0552 0.0497 0.0408 0.0363 0.0305
## Cumulative Proportion 0.688 0.7875 0.8427 0.8925 0.9332 0.9695 1.0000



##
## Call:
## lm(formula = uv ~ PC1 + PC2 + robot.group, data = pca)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.524 -0.767 -0.339 0.666 2.745
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.4086 0.0615 22.90 < 0.0000000000000002 ***
## PC1 0.0164 0.0119 1.37 0.17
## PC2 -0.2039 0.0315 -6.48 0.00000000013579 ***
## robot.grouphuman-like 0.6396 0.0869 7.36 0.00000000000034 ***
## robot.grouprobotic 0.4034 0.0698 5.78 0.00000000936594 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.923 on 1238 degrees of freedom
## Multiple R-squared: 0.0711, Adjusted R-squared: 0.0681
## F-statistic: 23.7 on 4 and 1238 DF, p-value: <0.0000000000000002
##
## Call:
## lm(formula = uv ~ PC1 + PC2 + robot.group, data = pca)
##
## Standardized Coefficients::
## (Intercept) PC1 PC2
## 0.00000 0.03761 -0.17778
## robot.grouphuman-like robot.grouprobotic
## 0.25810 0.20304
##
## Pearson's product-moment correlation
##
## data: pca$PC1 and pca$PC2
## t = -0.00000000000021, df = 1200, p-value = 1
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## -0.0556 0.0556
## sample estimates:
## cor
## -0.000000000000006098
##
## Call:
## lm(formula = uv ~ PC1 + PC2 + robot.group + Sex + Age, data = pca)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.554 -0.738 -0.353 0.611 2.756
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.34797 0.10507 12.83 < 0.0000000000000002 ***
## PC1 0.02238 0.01322 1.69 0.091 .
## PC2 -0.20486 0.03150 -6.50 0.00000000011393 ***
## robot.grouphuman-like 0.63961 0.08690 7.36 0.00000000000033 ***
## robot.grouprobotic 0.40417 0.06975 5.79 0.00000000867413 ***
## SexMale -0.06487 0.05271 -1.23 0.219
## Age 0.01194 0.00978 1.22 0.222
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.923 on 1236 degrees of freedom
## Multiple R-squared: 0.0735, Adjusted R-squared: 0.069
## F-statistic: 16.3 on 6 and 1236 DF, p-value: <0.0000000000000002
##
## Call:
## lm(formula = uv ~ PC1 + PC2 + robot.group + Sex + Age, data = pca)
##
## Standardized Coefficients::
## (Intercept) PC1 PC2
## 0.00000 0.05137 -0.17859
## robot.grouphuman-like robot.grouprobotic SexMale
## 0.25811 0.20342 -0.03383
## Age
## 0.03702
K-means clustering of a data-driven components

## [1] 1094
Distribution of robots among k-means clusters (PCA)
##
## 1 2 3
## actroid 1 0 0
## atlas 1 0 0
## kb 1 0 0
## kf 1 0 0
## spot 1 0 0
## nao 0 1 0
## pepper 0 1 0
## sofia 0 1 0
## stan 0 1 0
## tapia 0 1 0
## festo 0 0 1
##
## Call:
## lm(formula = uv ~ cluster, data = km)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.974 -0.798 -0.298 0.702 2.540
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.7977 0.0312 57.61 < 0.0000000000000002 ***
## cluster2 0.1761 0.0787 2.24 0.025 *
## cluster3 -0.3380 0.0837 -4.04 0.000057 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.948 on 1240 degrees of freedom
## Multiple R-squared: 0.0194, Adjusted R-squared: 0.0178
## F-statistic: 12.3 on 2 and 1240 DF, p-value: 0.0000053
##
## Call:
## lm(formula = uv ~ cluster, data = km)
##
## Standardized Coefficients::
## (Intercept) cluster2 cluster3
## 0.00000 0.06363 -0.11486
Unfolding analysis
Latent Class Analysis